ADR 036: add benchmarking workloads

[robsdedude]

Closes: DRIVERS-354

[AndyHeap-NeoTech]

Looks good, a possible bug to sanity check and one comment.

[Copilot]

Pull request overview

Adds an internal benchmarking runner and a suite of driver workloads under benchmarks/ (ADR-036 / DRIVERS-354), plus wiring for running/type-checking the benchmarks in development and Docker.

Changes:

• Introduce a bench benchmark runner (CLI, registry, timing, context) and workload modules for common serialization/deserialization + API scenarios.
• Add a dedicated benchmarks/ project (pyproject, lockfile, Dockerfile, README) to run benchmarks locally or in containers.
• Add ty to typing dependencies and a pre-commit hook to type-check the benchmarks project.

Reviewed changes

Copilot reviewed 50 out of 52 changed files in this pull request and generated 10 comments.

Show a summary per file

File	Description
pyproject.toml	Adds ty to the main repo typing dependency group.
benchmarks/uv.lock	Adds a frozen lockfile for the benchmarks environment.
benchmarks/src/bench/workloads/util/const.py	Defines integer boundary constants used by workloads.
benchmarks/src/bench/workloads/util/_timer.py	Implements timing collection + CSV flushing for benchmark results.
benchmarks/src/bench/workloads/util/_deserialization_template.py	Shared template for deserialization workloads (setup + timed work).
benchmarks/src/bench/workloads/util/_benchmark_registry.py	Registry + decorator for workload discovery and retry wrapper.
benchmarks/src/bench/workloads/util/_benchmark_context.py	Connection/auth context and driver factory for workloads.
benchmarks/src/bench/workloads/util/_bencher.py	Core benchmarking loop logic (warmup/timed iterations + retries).
benchmarks/src/bench/workloads/util/init.py	Exposes benchmark utility APIs for workloads/runner.
benchmarks/src/bench/workloads/serialize_vector.py	Workload: serialize vectors via execute_query.
benchmarks/src/bench/workloads/serialize_string.py	Workload: serialize large string payloads.
benchmarks/src/bench/workloads/serialize_null.py	Workload: serialize large null lists.
benchmarks/src/bench/workloads/serialize_list.py	Workload: serialize deeply nested lists.
benchmarks/src/bench/workloads/serialize_integer.py	Workload: serialize mixed-size integer payloads.
benchmarks/src/bench/workloads/serialize_float.py	Workload: serialize float payloads.
benchmarks/src/bench/workloads/serialize_dictionary.py	Workload: serialize nested dictionaries.
benchmarks/src/bench/workloads/serialize_bytes.py	Workload: serialize bytes payloads.
benchmarks/src/bench/workloads/serialize_boolean.py	Workload: serialize large boolean lists.
benchmarks/src/bench/workloads/record_streaming.py	Workload: stream a large number of records.
benchmarks/src/bench/workloads/home_db_resolution.py	Workload: measure home DB resolution behavior.
benchmarks/src/bench/workloads/deserialize_vector.py	Workload: deserialize vectors from properties.
benchmarks/src/bench/workloads/deserialize_string.py	Workload: deserialize large string property.
benchmarks/src/bench/workloads/deserialize_relationship.py	Workload: deserialize relationships + collections.
benchmarks/src/bench/workloads/deserialize_path.py	Workload: deserialize paths.
benchmarks/src/bench/workloads/deserialize_null.py	Workload: deserialize large list of nulls.
benchmarks/src/bench/workloads/deserialize_node.py	Workload: deserialize nodes from match results.
benchmarks/src/bench/workloads/deserialize_list.py	Workload: deserialize deeply nested list structure.
benchmarks/src/bench/workloads/deserialize_integer.py	Workload: deserialize mixed-size integer payloads.
benchmarks/src/bench/workloads/deserialize_float.py	Workload: deserialize float list property.
benchmarks/src/bench/workloads/deserialize_dictionary.py	Workload: deserialize nested map structures.
benchmarks/src/bench/workloads/deserialize_bytes.py	Workload: deserialize bytes property.
benchmarks/src/bench/workloads/deserialize_boolean.py	Workload: deserialize large boolean lists.
benchmarks/src/bench/workloads/cold_driver.py	Workload: measure cold driver creation overhead.
benchmarks/src/bench/workloads/chunking.py	Workload: measure large payload chunking behavior.
benchmarks/src/bench/workloads/api_unmanaged_transaction.py	Workload: unmanaged tx API usage.
benchmarks/src/bench/workloads/api_session_run.py	Workload: session.run API usage.
benchmarks/src/bench/workloads/api_managed_transaction.py	Workload: execute_write (managed tx) API usage.
benchmarks/src/bench/workloads/api_execute_query.py	Workload: execute_query API usage.
benchmarks/src/bench/workloads/_data_vector.py	Shared vector test data definitions.
benchmarks/src/bench/workloads/_data_float.py	Shared float test data definitions.
benchmarks/src/bench/workloads/_data_dictionary.py	Shared dictionary test data definitions.
benchmarks/src/bench/workloads/_data_bytes.py	Shared bytes test data definitions.
benchmarks/src/bench/workloads/init.py	Auto-imports workload modules so registry picks them up.
benchmarks/src/bench/main.py	Benchmark runner CLI entrypoint.
benchmarks/src/bench/init.py	Declares the bench package.
benchmarks/README.md	Documents local and Docker benchmark execution + lockfile maintenance.
benchmarks/pyproject.toml	Adds a dedicated benchmarks project configuration.
benchmarks/Dockerfile	Adds a Docker image to run benchmarks reproducibly.
benchmarks/.gitignore	Ignores benchmarks venv/results artifacts.
.pre-commit-config.yaml	Adds a ty check hook for the benchmarks project.
.gitignore	Ignores Ruff/Mypy cache directories.
.dockerignore	Updates build context exclusions (incl. benchmark results).

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 50 out of 52 changed files in this pull request and generated 1 comment.

[Copilot]

Pull request overview

Copilot reviewed 51 out of 53 changed files in this pull request and generated no new comments.

[AndyHeap-NeoTech]

Looks good. Thanks for the clarifications.